----------------------------------------------------------------------- 11 .37 .41 8.0 +65000 -1.8 1.2 27.50 0.429000 12 .31 .32 10.2 +34000 0.5 3.0 5.25 0.051300 13 .16 .16 20.4 -24000 7.9 8.9 0.023 0.000055 14 .18 .23 14.2 +69000 6.3 8.1 0.048 0.000238 15 .19 .... .... ...... .... 10.4 0.0057 ........

----------------------------------------------------------------------- 16 .41 .76 4.3 +20000 6.2 10.7 0.0044 0.000238 17 .19 .22 14.8 -20000 8.2 9.9 0.009 0.000041 18 .34 .... .... ...... .... 14.7 0.00011 ........

19 .19 .... .... ...... .... 9.9 0.009 ........

20 .76 1.03 3.2 -28000 -0.5 4.6 1.20 0.117500 ----------------------------------------------------------------------- 21 .17 .22 14.8 -598000 4.0 5.8 0.40 0.001815 22 .18 .19 17.1 -36000 5.6 7.1 0.12 0.000412 23 .18 .... .... ...... .... 9.7 0.011 ........

24 .19 .... .... ...... .... 7.1 0.12 ........

25 .17 .17 19.2 +21000 5.7 7.1 0.12 0.000329 ----------------------------------------------------------------------- 26 .22 .... .... ...... .... 10.8 0.004 ........

27 .53 .70 4.7 +10000 9.1 13.3 0.0025 0.000114 28 .19 .... .... ...... .... 5.7 0.44 ........

29 .29 .... .... ...... .... 11.1 0.0030 ........

30 .20 .23 14.2 -49000 4.5 6.3 0.25 0.001238 ----------------------------------------------------------------------- 31 .21 .51 6.4 +117000 -0.7 2.8 6.30 0.153600 32 .30 .38 8.6 +19000 5.1 8.0 0.053 0.000715 33 .25 .26 12.6 -11000 6.6 8.6 0.030 0.000189 34 .28 .31 10.5 +17000 4.6 7.0 0.13 0.001230 35 .26 .... .... ....... .... 11.3 0.0025 ........

----------------------------------------------------------------------- 36 .29 .29 11.2 -3000 7.1 9.4 0.014 0.000111 37 .17 .... .... ....... .... 9.9 0.009 ........

38 .22 .22 14.8 -7000 8.2 9.9 0.009 0.000041 -----------------------------------------------------------------------

On the basis of column 14 and of the movements and distances of the stars as given in the other columns Fig. 10 has been prepared. This gives an estimate of the approximate electrical energy received by the sun from the nearest stars for 70,000 years before and after the present. It is based on the twenty-six stars for which complete data are available in Table 6. The inclusion of the other twelve would not alter the form of the curve, for even the largest of them would not change any part by more than about half of 1 per cent, if as much. Nor would the curve be visibly altered by the omission of all except four of the twenty-six stars actually used. The four that are important, and their relative luminosity when nearest the sun, are Sirius 429,000, Altair 153,000, Alpha Centauri 117,500, and Procyon 51,300. The figure for the next star is only 4970, while for this star combined with the other twenty-one that are unimportant it is only 24,850.

Figure 10 is not carried more than 70,000 years into the past or into the future because the stars near the sun at more remote times are not included among the thirty-eight having the largest known parallaxes.

That is, they have either moved away or are not yet near enough to be included. Indeed, as Dr. Schlesinger strongly emphasizes, there may be swiftly moving, bright or gigantic stars which are now quite far away, but whose inclusion would alter Fig. 10 even within the limits of the 140,000 years there shown. It is almost certain, however, that the most that these would do would be to raise, but not obliterate, the minima on either side of the main maximum.

[Ill.u.s.tration: _Fig. 10. Climatic changes of 140,000 years as inferred from the stars._]

In preparing Fig. 10 it has been necessary to make allowance for double stars. Pa.s.sing by the twenty-two unimportant stars, it appears that the companion of Sirius is eight or ten magnitudes smaller than that star, while the companions of Procyon and Altair are five or more magnitudes smaller than their bright comrades. This means that the luminosity of the faint components is at most only 1 per cent of that of their bright companions and in the case of Sirius not a hundredth of 1 per cent.

Hence their inclusion would have no visible effect on Fig. 10. In Alpha Centauri, on the other hand, the two components are of almost the same magnitude. For this reason the effective radiation of that star as given in column 14 is doubled in Fig. 10, while for another reason it is raised still more. The other reason is that if our inferences as to the electrical effect of the sun on the earth and of the planets on the sun are correct, double stars, as we have seen, must be much more effective electrically than single stars. By the same reasoning two bright stars close together must excite one another much more than a bright star and a very faint one, even if the distances in both cases are the same. So, too, other things being equal, a triple star must be more excited electrically than a double star. Hence in preparing Fig. 10 all double stars receive double weight and each part of Alpha Centauri receives an additional 50 per cent because both parts are bright and because they have a third companion to help in exciting them.

According to the electro-stellar hypothesis, Alpha Centauri is more important climatically than any other star in the heavens not only because it is triple and bright, but because it is the nearest of all stars, and moves fairly rapidly. Sirius and Procyon move slowly in respect to the sun, only about eleven and eight kilometers per second respectively, and their distances at minimum are fairly large, that is, 8 and 10.2 light years. Hence their effect on the sun changes slowly.

Altair moves faster, about twenty-six kilometers per second, and its minimum distance is 6.4 light years, so that its effect changes fairly rapidly. Alpha Centauri moves about twenty-four kilometers per second, and its minimum distance is only 3.2 light years. Hence its effect changes very rapidly, the change in its apparent luminosity as seen from the sun amounting at maximum to about 30 per cent in 10,000 years against 14 per cent for Altair, 4 for Sirius, and 2 for Procyon. The vast majority of the stars change so much more slowly than even Procyon that their effect is almost uniform. All the stars at a distance of more than perhaps twenty or thirty light years may be regarded as sending to the sun a practically unchanging amount of radiation. It is the bright stars within this limit which are important, and their importance increases with their proximity, their speed of motion, and the brightness and number of their companions. Hence Alpha Centauri causes the main maximum in Fig. 10, while Sirius, Altair, and Procyon combine to cause a general rise of the curve from the past to the future.

Let us now interpret Fig. 10 geologically. The low position of the curve fifty to seventy thousand years ago suggests a mild inter-glacial climate distinctly less severe than that of the present. Geologists say that such was the case. The curve suggests a glacial epoch culminating about 28,000 years ago. The best authorities put the climax of the last glacial epoch between twenty-five and thirty thousand years ago. The curve shows an amelioration of climate since that time, although it suggests that there is still considerable severity. The retreat of the ice from North America and Europe, and its persistence in Greenland and Antarctica agree with this. And the curve indicates that the change of climate is still persisting, a conclusion in harmony with the evidence as to historic changes.

If Alpha Centauri is really so important, the effect of its variations, provided it has any, ought perhaps to be evident in the sun. The activity of the star's atmosphere presumably varies, for the orbits of the two components have an eccentricity of 0.51. Hence during their period of revolution, 81.2 years, the distance between them ranges from 1,100,000,000 to 3,300,000,000 miles. They were at a minimum distance in 1388, 1459, 1550, 1631, 1713, 1794, 1875, and will be again in 1956. In Fig. 11, showing sunspot variations, it is noticeable that the years 1794 and 1875 come just at the ends of periods of unusual solar activity, as indicated by the heavy horizontal line. A similar period of great activity seems to have begun about 1914. If its duration equals the average of its two predecessors, it will end about 1950. Back in the fourteenth century a period of excessive solar activity, which has already been described, culminated from 1370 to 1385, or just before the two parts of Alpha Centauri were at a minimum distance. Thus in three and perhaps four cases the sun has been unusually active during a time when the two parts of the star were most rapidly approaching each other and when their atmospheres were presumably most disturbed and their electrical emanations strongest.

[Ill.u.s.tration: _Fig. 11. Sunspot curve showing cycles, 1750 to 1920._

_Note._ The asterisks indicate two absolute minima of sunspots in 1810 and 1913, and the middle years (1780 and 1854) of two periods when the sunspot maxima never fell below 95. If Alpha Centauri has an effect on the sun's atmosphere, the end of another such period would be expected not far from 1957.]

The fact that Alpha Centauri, the star which would be expected most strongly to influence the sun, and hence the earth, was nearest the sun at the climax of the last glacial epoch, and that today the solar atmosphere is most active when the star is presumably most disturbed may be of no significance. It is given for what it is worth. Its importance lies not in the fact that it proves anything but that no contradiction is found when we test the electro-stellar hypothesis by facts which were not thought of when the hypothesis was framed. A vast amount of astronomical work is still needed before the matter can be brought to any definite conclusion. In case the hypothesis stands firm, it may be possible to use the stars as a help in determining the exact chronology of the later part of geological times. If the hypothesis is disproved, it will merely leave the question of solar variations where it is today.

It will not influence the main conclusions of this book as to the causes and nature of climatic changes. Its value lies in the fact that it calls attention to new lines of research.

FOOTNOTES:

[Footnote 120: Lewis Boss: Convergent of a Moving Cl.u.s.ter in Taurus; Astronom. Jour., Vol. 26, No. 4, 1908, pp. 31-36.]

[Footnote 121: F. R. Moulton: in Introduction to Astronomy, 1916.]

[Footnote 122: A. Penck: Die Alpen im Eiszeitalter, Leipzig, 1909.]

[Footnote 123: R. D. Salisbury: Physical Geography of the Pleistocene, in Outlines of Geologic History, by Willis and Salisbury, 1910, pp.

273-274.]

[Footnote 124: Davis, Pumpelly, and Huntington: Explorations in Turkestan, Carnegie Inst. of Wash., No. 26, 1905.

In North America the stages have been the subject of intensive studies on the part of Taylor, Leverett, Goldthwait, and many others.]

CHAPTER XVI

THE EARTH'S CRUST AND THE SUN

Although the problems of this book may lead far afield, they ultimately bring us back to the earth and to the present. Several times in the preceding pages there has been mention of the fact that periods of extreme climatic fluctuations are closely a.s.sociated with great movements of the earth's crust whereby mountains are uplifted and continents upheaved. In attempting to explain this a.s.sociation the general tendency has been to look largely at the past instead of the present. Hence it has been almost impossible to choose among three possibilities, all beset with difficulties. First, the movements of the crust may have caused the climatic fluctuations; second, climatic changes may cause crustal movements; and third, variations in solar activity or in some other outside agency may give rise to both types of terrestrial phenomena.

The idea that movements of the earth's crust are the main cause of geological changes of climate is becoming increasingly untenable as the complexity and rapidity of climatic changes become more clear, especially during post-glacial times. It implies that the earth's surface moves up and down with a speed and facility which appear to be out of the question. If volcanic activity be invoked the problem becomes no clearer. Even if volcanic dust should fill the air frequently and completely, neither its presence nor absence would produce such peculiar features as the localization of glaciers, the distribution of loess, and the mild climate of most parts of geological time. Nevertheless, because of the great difficulties presented by the other two possibilities many geologists still hold that directly or indirectly the greater climatic changes have been mainly due to movements of the earth's crust and to the reaction of the crustal movements on the atmosphere.

The possibility that climatic changes are in themselves a cause of movements of the earth's crust seems so improbable that no one appears to have investigated it with any seriousness. Nevertheless, it is worth while to raise the question whether climatic extremes may cooperate with other agencies in setting the time when the earth's crust shall be deformed.

As to the third possibility, it is perfectly logical to ascribe both climatic changes and crustal deformation to some outside agency, solar or otherwise, but hitherto there has been so little evidence on this point that such an ascription has merely begged the question. If heavenly bodies should approach the earth closely enough so that their gravitational stresses caused crustal deformation, all life would presumably be destroyed. As to the sun, there has. .h.i.therto been no conclusive evidence that it is related to crustal movements, although various writers have made suggestions along this line. In this chapter we shall carry these suggestions further and shall see that they are at least worthy of study.

As a preliminary to this study it may be well to note that the coincidence between movements of the earth's crust and climatic changes is not so absolute as is sometimes supposed. For example, the profound crustal changes at the end of the Mesozoic were not accompanied by widespread glaciation so far as is yet known, although the temperature appears to have been lowered. Nor was the violent volcanic and diastrophic activity in the Miocene a.s.sociated with extreme climates.

Indeed, there appears to have been little contrast from zone to zone, for figs, bread fruit trees, tree ferns, and other plants of low lat.i.tudes grew in Greenland. Nevertheless, both at the end of the Mesozoic and in the Miocene the climate may possibly have been severe for a time, although the record is lost. On the other hand, Kirk's recent discovery of glacial till in Alaska between beds carrying an undoubted Middle Silurian fauna indicates glaciation at a time when there was little movement of the crust so far as yet appears.[125] Thus we conclude that while climatic changes and crustal movements usually occur together, they may occur separately.

According to the solar-cyclonic hypothesis such a condition is to be expected. If the sun were especially active when the terrestrial conditions prohibited glaciation, changes of climate would still occur, but they would be milder than under other circ.u.mstances, and would leave little record in the rocks. Or there might be glaciation in high lat.i.tudes, such as that of southern Alaska in the Middle Silurian, and none elsewhere. On the other hand, when the sun was so inactive that no great storminess occurred, the upheaval of continents and the building of mountains might go on without the formation of ice sheets, as apparently happened at the end of the Mesozoic. The lack of absolute coincidence between glaciation and periods of widespread emergence of the lands is evident even today, for there is no reason to suppose that the lands are notably lower or less extensive now than they were during the Pleistocene glaciation. In fact, there is much evidence that many areas have risen since that time. Yet glaciation is now far less extensive than in the Pleistocene. Any attempt to explain this difference on the basis of terrestrial changes is extremely difficult, for the shape and alt.i.tude of continents and mountains have not changed much in twenty or thirty thousand years. Yet the present moderately mild epoch, like the puzzling inter-glacial epochs of earlier times, is easily explicable on the a.s.sumption that the sun's atmosphere may sometimes vary in harmony with crustal activity, but does not necessarily do so at all times.

Turning now to the main problem of how climatic changes may be connected with movements of the earth's crust, let us follow our usual method and examine what is happening today. Let us first inquire whether earthquakes, which are one of the chief evidences that crustal movements are actually taking place in our own times, show any connection with sunspots. In order to test this, we have compared _Milne's Catalogue of Destructive Earthquakes_ from 1800 to 1899, with Wolf's sunspot numbers for the same period month by month. The earthquake catalogue, as its compiler describes it, "is an attempt to give a list of earthquakes which have announced changes of geological importance in the earth's crust; movements which have probably resulted in the creation or the extension of a line of fault, the vibrations accompanying which could, with proper instruments, have been recorded over a continent or the whole surface of our world. Small earthquakes have been excluded, while the number of large earthquakes both for ancient and modern times has been extended. As an ill.u.s.tration of exclusion, I may mention that between 1800 and 1808, which are years taken at random, I find in Mallet's catalogue 407 entries. Only thirty-seven of these, which were accompanied by structural damage, have been retained. Other catalogues such as those of Perry and Fuchs have been treated similarly."[126]

If the earthquakes in such a carefully selected list bear a distinct relation to sunspots, it is at least possible and perhaps probable that a similar relation may exist between solar activity and geological changes in the earth's crust. The result of the comparison of earthquakes and sunspots is shown in Table 7. The first column gives the sunspot numbers; the second, the number of months that had the respective spot numbers during the century from 1800 to 1899. Column C shows the total number of earthquakes during the months having any particular degree of spottedness; while D, which is the significant column, gives the average number of destructive earthquakes per month under each of the six conditions of solar spottedness. The regularity of column D is so great as to make it almost certain that we are here dealing with a real relationship. Column F, which shows the average number of earthquakes in the month succeeding any given condition of the sun, is still more regular except for the last entry.

TABLE 7

DESTRUCTIVE EARTHQUAKES FROM 1800 TO 1899 COMPARED WITH SUNSPOTS

A: _Sunspot numbers_ B: _Number of months per Wolf's Table_ C: _Number of earthquakes_ D: _Average number of earthquakes per month_ E: _Number of earthquakes in succeeding month_ F: _Average number of earthquakes in succeeding month_

A B C D E F

0-15 344 522 1.52 512 1.49 15-30 194 306 1.58 310 1.60 30-50 237 433 1.83 439 1.85 50-70 195 402 2.06 390 2.00 70-100 135 286 2.12 310 2.30 over 100 95 218 2.30 175 1.84

The chance that six numbers taken at random will arrange themselves in any given order is one in 720. In other words, there is one chance in 720 that the regularity of column D is accidental. But column F is as regular as column D except for the last entry. If columns D and E were independent there would be one chance in about 500,000 that the six numbers in both columns would fall in the same order, and one chance in 14,400 that five numbers in each would fall in the same order. But the two columns are somewhat related, for although the after-shocks of a great earthquake are never included in Milne's table, a world-shaking earthquake in one region during a given month probably creates conditions that favor similar earthquakes elsewhere during the next month. Hence the probability that we are dealing with a purely accidental arrangement in Table 7 is less than one in 14,400 and greater than one in 500,000. It may be one in 20,000 or 100,000. In any event it is so slight that there is high probability that directly or indirectly sunspots and earthquakes are somehow connected.

In ascertaining the relation between sunspots and earthquakes it would be well if we could employ the strict method of correlation coefficients. This, however, is impossible for the entire century, for the record is by no means h.o.m.ogeneous. The earlier decades are represented by only about one-fourth as many earthquakes as the later ones, a condition which is presumably due to lack of information. This makes no difference with the method employed in Table 7, since years with many and few sunspots are distributed almost equally throughout the entire nineteenth century, but it renders the method of correlation coefficients inapplicable. During the period from 1850 onward the record is much more nearly h.o.m.ogeneous, though not completely so. Even in these later decades, however, allowance must be made for the fact that there are more earthquakes in winter than in summer, the average number per month for the fifty years being as follows:

Jan. 2.8 May 2.4 Sept. 2.5 Feb. 2.4 June 2.3 Oct. 2.6 Mar. 2.5 July 2.4 Nov. 2.7 Apr. 2.4 Aug. 2.4 Dec. 2.8

The correlation coefficient between the departures from these monthly averages and the corresponding departures from the monthly averages of the sunspots for the same period, 1850-1899, are as follows:

Sunspots and earthquakes of same month: +0.042, or 1.5 times the probable error.

Sunspots of a given month and earthquakes of that month and the next: +0.084, or 3.1 times the probable error.

Sunspots of three consecutive months and earthquakes of three consecutive months allowing a lag of one month, i.e., sunspots of January, February, and March compared with earthquakes of February, March, and April; sunspots of February, March, and April with earthquakes of March, April, and May, etc.; +0.112, or 4.1 times the probable error.